KR20070038881A - Method and system for storing data packets - Google Patents

Abstract

The present invention provides a method for managing the storage of data packets, each of which consists of at least two data blocks, in a storage space of a buffer, the method comprising: receiving (40) data blocks of a current data packet; Writing the data blocks of the received current data packet to the free storage space of the buffer (42). If there is not enough free space in the buffer to write the received data block of the current data packet, the method stops writing the received data blocks of the current data packet (48), and at least one write of the current data packet. Defines the completed data block as part of the free storage space, so that the received data blocks of the next data packet can be overwritten by the overwritten data blocks of the current data packet.

Applied to digital TV reception.

Data packet, buffer, free storage, read, write, overwrite

Description

Method and system for storing data packets

1 illustrates a data packet.

2 illustrates a system for storing data packets, such as the data packet of FIG.

3 illustrates the system of FIG. 2 in an overflow situation;

4A and 4B are block diagrams illustrating a method of managing storage of data packets.

<Explanation of symbols for the main parts of the drawings>

10. System 12. Buffer

14. Control means 16. Read pointer management unit

18. Write Pointer Manager R. Read Pointer

W. Write Pointer

Technical Field of the Invention

The present invention relates to a method for managing the storage of data packets in a storage of a buffer.

Background of the Invention

Full motion digital television requires large amounts of storage and data transmission bandwidth. Thus, digital television systems use video compression algorithms to reduce the amount of storage and data transfer bandwidth required. Current digital television systems use MPEG or MPEG-2 encoding to reduce the data transmission bandwidth required. MPEG-2 is defined in ITU-T H.222.0 (ISO / IEC 13818-1) as "Generic Coding of Moving Pictures And Associated Audio Information." Thus, visual auditory sequences, such as movies or other content, are stored on the media server in MPEG-2 compression format, and the visual auditory sequences are decoded and generally converted to analog format for display and / or delivery to the user.

It is well known that a transport stream (TS) according to MPEG-2 is a stream including compressed digital video, compressed digital audio, and the like. The transport stream is used in digital broadcasting and includes a plurality of programs in multiple modes.

Two kinds of MPEG-2 receivers are used. First, in set top boxes, input MPEG-2 TS packets are handled by dedicated hardware functionality, namely MPEG-2 TS demultiplexers, which can handle input bit streams. In this case, there is no data overflow. In addition, if exceptional cases such as temporary loss of signal occur, the MPEG-2 TS reader may be easily resynchronized by using dedicated hardware functions.

The second type of MPEG-2 receiver uses a PC using PCI or USB interfaces. Generally, PCI MPEG2-TS receivers are PCI masters and FIFO overflow does not occur because MPEG2-TS packets are written (written) to PC system memory by dedicated hardware. In addition, USB MPEG2-TS receivers use isochronous USB endpoints to transmit MPEG2-TS packets, and with the major drawback of this type of transmission, which retains USB bandwidth by the actual bit rate. Guaranteed read bit rate.

Therefore, all these MPEG-2 receivers do not experience overflow at the expense of high complexity.

On the other hand, there are simpler USB MPEG-2 TS receivers that do not use USB isochronous endpoints.

However, regardless of satellite, cable or terrestrial transmission modes, problems of data overflow and overflow management can occur in these receivers. Conventionally, overflow management has not guaranteed a higher data read bit rate than the received data bit rate (e.g., a PC that may be temporarily overloaded) and maintains the synchronization required for MPEG-2 TS packets. Can not. In particular, the MPEG-2 TS reader is capable of receiving full MPEG-2 TS packets and is vulnerable to ensuring that it remains synchronized if the data read bit rate is temporarily lower than the received bit rate.

Therefore, there is a need for an improved method and system for storing data packets and managing overflow situations in a simple and economical way.

Summary of the Invention

The present invention therefore provides an improved method for storing data packets according to claim 1. Other features of this method are described in claims 2 to 5 which are dependent claims.

The invention further provides a system for storing data packets according to claims 6 or 7.

The invention further provides an MPEG-2 TS receiver according to claim 8 and a computer program product according to claim 9.

details

1 shows a data packet DP composed of four data blocks db1, db2, db3, and db4. In this embodiment, the data packet DP is a stream packet for MPEG-2 transmission. Each data block db1, db2, db3, db4 contains data A, B, C, and D, respectively.

FIG. 2 shows a system 10 for storing data packets, such as the data packet DP of FIG. 1. The system 10 is a first-in-first-out buffer 12 having storage space and management means 14 for managing the storage of data packets in the storage space of the buffer 12. It includes. In the description below, both the buffer and the storage space of the buffer are described by reference numeral 12.

The buffer 12 includes eight data cells dc1 to dc8 which are to store data. Each is associated with a unique address. The buffer 12 may be executed in a random access memory (RAM) of a computer.

In the present description, the data cells and data blocks have the same size, for example 1 byte. Of course, they can have different sizes.

The management means 14 includes a read pointer management unit 16 and a write pointer management unit 18.

The read pointer manager 16 manages a read pointer R indicating a position in the buffer storage space 12 of the data cell to which data is to be read. In FIG. 2, the read pointer R includes the address of the data cell dc4.

The write pointer manager 18 manages a write pointer W indicating a position in the buffer storage space 12 of the data cell to which data is to be written. In FIG. 2, the write pointer W includes the address of the data cell dc6.

Both pointers R and W travel on the buffer storage space along the same proceeding loop path from data cell dc1 to data cell dc8 and back to data cell dc1.

The management means 14 may also store the address of the data cell in which the first data block of the data packet is recorded. This will be explained below.

Preferably, read / write pointers R and W are stored in a pointer register (not shown).

3 illustrates an overflow situation, ie there is not enough free storage space to write a received data block. Free storage is either data cells without data or data cells from which data has already been read. Thus, the free storage space of the buffer 12 is defined by the " distance " from the write pointer W to the read pointer R on the progress path, that is, the write pointer W and the read pointer R. It is composed by the data cells of the liver.

In the overflow situation of FIG. 3, the write pointer W follows the read pointer R on the progress path. Thus, all storage space in the buffer is filled with unread data.

In the example described, the preceding three data blocks db1, db2, and db3 are received and their corresponding data A, B, C are written. However, it is impossible to record the data D of the received data block db4.

4A and 4B are block diagrams of a method for managing the storage of data packets in the storage space of the buffer 12, which is achieved by the management means 14.

Management of storing data packets includes a data packet reading method shown in FIG. 4A and a data packet writing method shown in FIG. 4B. These two methods are executed in parallel.

Referring to FIG. 4A, the read method includes a first step 30 of reading data written to a data cell indicated by a read pointer R. Referring to FIG.

The read method then proceeds to step 32, where the readable data is written to the next data cell. If that is not the case, i.e. if the write pointer is pointing to the next data cell, the writing of new data is awaited, which is indicated by an arrow 34 back to test step 32.

If there is readable data, ie if the write pointer does not point to the next data cell, then the read method increments the read pointer R so that the read pointer R points to the next data cell (36). Proceeds back to the reading step 30 again.

     Referring to FIG. 4B, the write method includes a first step 40 of receiving a data block of a current data packet being received.

The write method then includes testing 42 if there is sufficient free storage space for writing a new block of data received later. This testing step is accomplished by determining storage space on the progressive annular path from the write pointer to the read pointer.

If there is sufficient free storage space, i.e., if the write pointer W is not equal to the read pointer R, the write method writes the received data block to the data cell indicated by the write pointer W, and writes. Proceeding to step 44, the write pointer W is incremented such that the pointer W points to the next data cell of the progressive annular path.

If the written data block is the first block of the data packet, for example, if it is the data block db1 of the data packet DP of Fig. 1, the position address of the data block is stored by the management means 14; do. This step is represented by block 46 of FIG. 4B. The write method then returns to the receiving step 40 for managing the storage of the received data blocks later in the current data packet.

If there is not enough free storage space, i.e., if the write pointer W is equal to the read pointer R, the write method stops writing the received data blocks of the current data packet and positions the write pointer W. The process proceeds to step 48 where the address is moved to the location address of the first written data block of the currently received data packet stored in step 46.

Next, in the waiting step 50, newly received data blocks of the current data packet are discarded, that is, not written to the buffer 12. Upon receiving the first block of new data packets, the method returns to receiving step 40.

By moving the write pointer W, the received data blocks of the next data packet are advantageously able to overwrite the already written data blocks of the previous data packet.

Indeed, in many applications, such as MPEG-2 data reception, incomplete data packets cannot be used and are discarded by the application using buffer 12.

Overwriting this data packet may further provide free storage for future data packets, thus reducing the number of times the overflow occurs again.

It provides an improved method and system for storing data packets and managing overflow situations in a simple and economical way.

Claims (9)

As a method for managing the storage of data packets each consisting of at least two data blocks (db1, db2, db3, db4) in the storage space of the buffer 12, Receiving (40) data blocks of a current data packet; A data storage management method comprising: writing data blocks of a received current data packet to a free storage space of the buffer 12; If there is not enough free space in the buffer 12 to write the received data block of the current data packet, stopping (48) writing the received data blocks of the current data packet; And Defining the at least one write block of the current data packet as part of the free storage space so that received data blocks of the next data packet can be overwritten with the write data blocks of the current data packet ( A data packet storage management method, characterized in that it comprises a. The method of claim 1, During the definition step (48), all data blocks for which writing of the current data packet is completed are defined as part of free storage space. The method according to claim 1 or 2, And the buffer (12) is a first-in-first-out (FIFO) buffer. The method of claim 3, wherein The write pointer W indicates the position in the buffer storage space 12 in which the data block is to be written, and the read pointer P indicates the position in the buffer storage space 12 in which the received write data block is to be read. The pointers travel on buffer storage along the same progressive annular path, the method comprising storing 43 the location of the first written data block of the current data packet, Determining 44 whether there is remaining free storage space 12 includes determining storage space 12 on the annular path from the write pointer W to the read pointer P, Defining at least one written block of data in the current data packet as part of free storage comprises moving the write pointer W to a stored location of the first write block of data in the current data packet. Data packet storage management method, characterized in that. The method according to any one of claims 1 to 4, Data packet storage management method, characterized in that the current data packet is a stream packet for MPEG-2 transmission. A system for storing data packets (DP) comprising a buffer having a storage space, each consisting of at least two data blocks (db1, db2, db3, db4), the system comprising: Management means (14) for managing the storage of data packets in a storage space, said management means (14) being arranged to carry out the method according to any one of the preceding claims. The method of claim 6, The buffer is a FIFO buffer, wherein said management means (14) is arranged to carry out the method according to claim 3 or 4. As an MPEG-2 receiver, An MPEG-2 receiver comprising a system according to claim 6 or 7 for storing stream data packets for MPEG-2 transmission. A computer program product mountable directly into an internal memory of a digital computer, the product comprising software code portions for performing the steps of any of claims 1 to 5 when the product is run on a computer. Program product.

Images